Imagine someone could know exactly how you're sitting in your chair right now, how many times you breathe per minute, and whether your heart is beating faster than normal. No cameras. No microphones. Without you having any clue it's happening. Just using the WiFi signals already passing through your walls.
That's WiFi DensePose, and it's not science fiction. It's open source, available right now, and the hardware you need costs about eight dollars.
The project, distributed under MIT license with SHA-256 integrity verification, achieves something that seemed reserved for labs with enormous budgets: mapping body poses in real time, detecting heart rate, and measuring respiratory frequency through solid walls. WiFi signals distort when they encounter moving objects, and those distortions, processed through trained neural networks, reveal information we normally associate with specialized medical equipment. What distinguishes it from similar projects at MIT and Carnegie Mellon is the combination of hardware accessibility and complete code openness. In my experience reviewing tech tools, this accessibility changes the entire landscape.
The project presents itself as "privacy-first." The argument is intuitive: without video there are no compromising images, no facial recognition, no visual material that could leak or be used to identify you. And that argument has some truth to it. But there's a paradox here. A system that detects your body, your breathing, and your pulse through walls isn't private just because it lacks a lens. It's invisible surveillance. The difference between a security camera and WiFi DensePose isn't that one monitors and the other doesn't. The difference is that you can see the camera.
This connects directly to Chapter 22 of the Luddite Manifesto, where I explore symmetric surveillance. The problem with monitoring systems isn't the technology itself, but the asymmetry of who can see whom. When a government or corporation deploys surveillance systems, there's a unidirectional relationship: they see you, you don't see them, and you rarely have access to what they do with what they collect. That asymmetry is the core issue. WiFi DensePose, in the hands of an unaccountable state, is no more private than a camera network. It's harder to detect, which makes it potentially riskier.
What does transform the landscape is that this system is completely auditable. Anyone with technical knowledge can review the code, verify that the SHA-256 hash corresponds to the original repository, and confirm there are no hidden functions sending data to external servers. This radical transparency is precisely what the Manifesto proposes as standard for technologies that touch personal data. It contrasts with corporate and government surveillance systems that operate as black boxes: PRISM, credit scoring algorithms, facial recognition systems used by police in dozens of countries without auditing their error rates. Code openness doesn't guarantee ethical use, but it ensures it can't hide what it does. Think of it as an open book versus a padlock.
The project's applications listing illustrates a pattern I've seen repeated throughout twelve thousand years of human history, from archaeological evidence in ancient settlements to modern tools. Every tool for social coordination ends up becoming a tool for control as well. The developers mention uses like monitoring elderly people living alone, fall detection without invading their visual privacy, disaster rescue where sensors detect survivors under rubble, and sleep pattern analysis for medicine. All valuable. And in the same document: military surveillance, border control, security facility monitoring. It's not hypocrisy. It's the nature of tools. Fire cooked the first meal and burned villages. Writing preserved knowledge and enabled tax records. The internet connected the world and built the most sophisticated surveillance infrastructure of our species.
The question has never been whether a tool will have dual uses. It always does. The question is what mechanisms exist to limit abuse, and that's not a technical answer, but a political and institutional one. Sometimes I wonder if in contexts like Mexico, where power dynamics are so visible, these tools could be used in community ways that balance control.
From the Manifesto's perspective, this poses an exercise: if the governance system I propose existed, how would WiFi DensePose be regulated? I don't have a definitive answer. It would be irresponsible to pretend otherwise. But I offer an experimental framework. Rotating committees for technological decisions would approve deployments in shared spaces, with limited mandates to prevent perpetuation in office. Auditable AI would verify that data is used only for declared purposes, with alerts for deviations. And symmetric surveillance would ensure that whoever deploys the system is also monitored: administrators who watch employees would accept auditable logging of their own behavior.
Would it work perfectly? No. No system does. But it generates better incentives than current ones, where surveillance is unaccountable. There are emerging alternatives, like democratic data trusts, that suggest paths toward greater equity.
Banning WiFi DensePose isn't realistic. The physics behind it won't disappear, and the knowledge is already globally distributed. The code exists in accessible repositories. Suppressing it would push development toward less transparent actors, while states and corporations would use it anyway. What's possible is building accountability frameworks as sophisticated as the technology. Researchers have long argued that privacy requires social structures that make abuse costly. Europe's GDPR, despite its flaws, is one step. Proposals for data trusts where citizens delegate control collectively are another. None solve everything, but they point to solutions in power relations, not just in technique.
WiFi DensePose is a clear mirror. Its openness makes it auditable in ways impossible for closed systems. Its low cost enables benign community uses. Its invisibility makes it risky without controls. All three truths coexist, and ignoring any one is incomplete.
The technology is here. The question is whether we build institutions with the same speed and creativity. Historically we arrive late, but there's room to change that. I write because I believe there are viable alternatives.
Stones don't lie, but historians sometimes do.
Sources:
1. WiFi-DensePose Repository — GitHub (MIT License, verifiable with SHA-256)
2. Zhao, M. et al. "Through-Wall Human Pose Estimation Using Radio Signals" — MIT CSAIL, 2018
3. Zhu, D. et al. "DensePose From WiFi" — Carnegie Mellon University, 2022
4. Solove, D. J. Understanding Privacy — Harvard University Press, 2008
5. Schneier, B. Data and Goliath: The Hidden Battles to Collect Your Data and Control Your World — W.W. Norton, 2015
